The triterpene glycoside glycyrrhizin is the main active compound in liquorice. It is used as a herbal medicine owing to its anticancer, antiviral and anti-inflammatory properties. Its mode of action, however, remains widely unknown. In the present study, we aimed to elucidate the molecular mechanism of glycyrrhizin in attenuating inflammatory responses in macrophages. Using microarray analysis, we found that glycyrrhizin caused a broad block in the induction of pro-inflammatory mediators induced by the TLR (Toll-like receptor) 9 agonist CpG-DNA in RAW 264.7 cells. Furthermore, we found that glycyrrhizin also strongly attenuated inflammatory responses induced by TLR3 and TLR4 ligands. The inhibition was accompanied by decreased activation not only of the NF-κB (nuclear factor κB) pathway but also of the parallel MAPK (mitogen-activated protein kinase) signalling cascade upon stimulation with TLR9 and TLR4 agonists. Further analysis of upstream events revealed that glycyrrhizin treatment decreased cellular attachment and/or uptake of CpG-DNA and strongly impaired TLR4 internalization. Moreover, we found that the anti-inflammatory effects were specific for membrane-dependent receptor-mediated stimuli, as glycyrrhizin was ineffective in blocking Tnfa (tumour necrosis factor α gene) induction upon stimulation with PMA, a receptor- and membrane-independent stimulus. These observations suggest that the broad anti-inflammatory activity of glycyrrhizin is mediated by the interaction with the lipid bilayer, thereby attenuating receptor-mediated signalling.
- cell membrane
- Toll-like receptor (TLR)
The innate immune system responds to invading microbes by detecting highly conserved structures of pathogens through interaction with TLRs (Toll-like receptors). In humans, ten TLRs (TLR1–TLR10) have been identified to date, detecting various pathogen-specific molecules [1,2]. TLR9 is activated by unmethylated CpG motifs within single-stranded DNA which are present in high amounts in bacteria. These motives can be mimicked by synthetic ODNs (oligodeoxynucleotides) [3–5]. In unstimulated cells, TLR9 is localized in the endoplasmatic reticulum. Upon activation, it translocates to the lysosomal compartment, where signalling is initiated [3,6,7]. TLR3 also localizes in endosomal compartments [8,9]. It can be activated by dsRNA (double-stranded RNA) derived from the viral RNA genome or produced during viral replication. To mimic viral infection, poly(I:C) (polyriboinosinic:polyribocytidylic acid), a stable synthetic dsRNA analogue can be used . TLR3 activation results in induction of several pro-inflammatory cytokines and the expression of type I IFNs (interferons), which in turn activate numerous antiviral proteins including Mx GTPase, protein kinase R, and 2′-5′-oligoadenylate synthetases . TLR4, which is expressed on the cell surface, can be activated by several bacterial surface molecules, including lipopeptides and LPS (lipopolysaccharide) from Gram-negative bacteria. Activation of TLR4 triggers strong induction of pro-inflammatory responses [11–14].
TLR4 activation depends on intact plasma membrane microdomains . They aid TLR4 signalling initiation by facilitating the formation of the active receptor complex consisting of TLR4–MD2 (myeloid differentiation factor 2) and CD14 . The cellular uptake of CpG-DNA mediated by scavenger receptors is similarly membrane-dependent [16–19]. In addition, membrane microdomains have been demonstrated to play a critical role in the initiation of a broad variety of signalling events, including activation of T- and B-cell receptor signalling. They act through facilitating the pre-assembly and stabilization of receptor complexes, which permits rapid, efficient and sustained connection to the signalling cascade upon receptor engagement [20,21]. Accordingly, disruption of plasma membrane microdomains results in strong attenuation of receptor-mediated signalling [15,20,21].
Activation of TLRs triggers the formation of distinct signalling complexes. Stimulus-specific responses are achieved by the recruitment of different adopter proteins. The activation of these pathways leads to the expression of various pro-inflammatory cytokines and chemokines such as TNFα (tumour necrosis factor α), IL (interleukin)-1β and RANTES (regulated upon activation, normal T-cell expressed and secreted), the secretion of IFNs and to apoptotic responses. The induction of these responses are mainly triggered by the activation of NF-κB (nuclear factor κB) and MAPK (mitogen-activated protein kinase) pathways, a consequence common to the stimulation of all TLRs [2,3,14].
Liquorice root has been used in traditional Chinese, Indian and Greek medicine since the Middle Ages [22,23]. It contains several compounds including triterpene saponins, flavonoids and polysaccharides . Of these, the triterpene glycoside glycyrrhizin has been reported to be beneficial against viral infections. In fact, the glycyrrhizin formulation SNMC (Stronger Neo-Minophagen C) has been used to treat viral hepatitis C in Japan [25,26]. Antiviral activity has also been reported against influenza, SARS (severe acute respiratory syndrome) and HIV-1 [27–30]. Apart from its antiviral activity, glycyrrhizin has been shown to influence inflammatory responses by up-regulating IL-10, IL-2, IL-12 and IFNγ, while down-regulating pro-inflammatory cytokines, including IL-8, IL-1β, TNFα and eotaxin-1, as well as COX (cyclo-oxygenase) activity in different experimental systems [31–37].
It is speculated that, owing to its structure, glycyrrhizin might act as glucocorticoid receptor agonist , whereas another study demonstrated a glucocorticoid receptor-independent block of NF-κB transcriptional activation . In fact, inhibition of the NF-κB activity by glycyrrhizin and other triterpenoids has been suggested for quite some time [32,39–42]. Interestingly, a study by Harada  reported the antiviral activity to be a result of changes in membrane fluidity possibly caused by incorporation of glycyrrhizin into the cellular membrane. However, the exact mode of action remains unclear.
In the present study, we aimed to elucidate the mode of action of glycyrrhizin as inhibitor of inflammatory responses mediated by TLRs. Initially, the effect of glycyrrhizin on the expression profile of CpG-DNA stimulated macrophages was characterized followed by an in-depth evaluation of the most significantly affected proteins. We extended these experiments further by analysing the effects of glycyrrhizin on inflammatory responses caused by TLR3 and TLR4 agonists. Furthermore, its inhibitory effect on NF-κB and MAPK activation was investigated. Finally, the effects of glycyrrhizin on membrane-dependent ligand binding/uptake and signalling were evaluated.
Cell culture and stimulations
RAW 264.7 macrophages (obtained from the A.T.C.C., Manassas, VA, U.S.A.) were cultured in DMEM (Dulbecco's modified Eagle's medium), with high glucose (Gibco) supplemented with 10% (v/v) FBS (fetal bovine serum) and penicillin/steptomycin (PAA) in a humidified atmosphere of 5% CO2. For stimulations, cells were seeded at a density of 2×105–106 cells/ml in DMEM supplemented with 5% (v/v) FBS and stimulated with 10 or 100 nM CpG-DNA (CpG-ODN1826; InvivoGen), 1, 10 or 2000 ng/ml LPS (Sigma), 1 μg/ml poly(I:C) (InvivoGen) and 10 or 100 nM PMA (Sigma).
Cells were stimulated with 10 nM CpG-DNA in the presence or absence of glycyrrhizin (1 mM) using biological duplicates. Alternatively, cells were treated with glycyrrhizin without further CpG-DNA stimulation. RNA was extracted 4 h after stimulation using the RNeasy mini kit (Qiagen) according to the manufacturer's instructions. The yield and purity of the RNA were determined using a ND-1000 spectrophotometer (NanoDrop Technologies). To generate fluorescently labelled cRNA targets for hybridization with mouse (4×44k) oligonucleotide arrays (G4122F), the Agilent Low RNA Input Fluorescent Linear Amplification kit was used. A 500 ng sample of total RNA was used for reverse transcription, the subsequent T7 polymerase cRNA target synthesis and for the labelling reaction with either Cy3 (indocarbocyanine)–CTP or Cy5 (indodicarbocyanine)–CTP nucleotides. Each of the biological duplicates was processed in duplicate reactions. RNeasy mini-spin columns (Qiagen) were used for purifying labelled cRNA targets. The cRNA was digested, and hybridizations were performed in a rotisserie at 65 °C for 17 h using Agilent's hybridization kit.
Microarray data analysis
Slides were scanned with the dual-laser Agilent microarray scanner at 5 μ-pixel resolution and 10 and 100% PMT (photo multiplier tube) power setting to allow XDR (extended dynamic range) data calculation. For raw data generation and normalization, Feature Extraction software (version 9.1; Agilent) was used. The Linear and LOWESS algorithm was used with default settings for signal characteristics and normalization spots. To correct for intensity variations at different regions on the array, spatial and multiplicative detrending (removing the surface trend) was selected (default). Feature Extraction output files were imported into Genespring GX version 9.0.2 (Agilent) for visualization and analysis of expression data. Data filtering according to a fold change criterion and a certain confidence level (P value) was completed to identify targets that were statistically significant. Where indicated, BH (Benjamini–Hochberg) correction was performed. GO (Gene Ontology) analyses was made with Genespring software.
Quantification of gene expression
mRNA expression levels of stimulated and unstimulated cells with or without glycyrrhizin treatment were quantified using QuantiGene technology (Panomics). Gene-specific probesets were obtained from Panomics, and luminescence was detected using a Synergy HT microplate reader (Biotek). Analysis was performed according to the manufacturer's instructions. Briefly, cells were lysed in the lysis buffer provided (Panomics) supplemented with proteinase K, by boiling at 60 °C for 30 min. Supernatants were stored at −80 °C until further analysis. Results were normalized to housekeeping genes [GAPDH (glyceraldehyde-3-phosphate dehydrogenase) or actin], which were analysed in parallel. Fold change was calculated in relation to unstimulated cells, which were set to 1.
The amount of cytokine release was determined, using ELISA, in culture supernatants of cells stimulated with the corresponding ligand overnight. TNFα and IL-1β ELISA kits were obtained from Biolegend, RANTES and MIP-1β (macrophage inflammatory protein 1β) ELISA kits were from R&D, IFNβ ELISA was from PBL and IL-6 ELISA was from Bender, and all ELISAs were carried out according to the manufacturer's instructions. Absorbance was detected using a Synergy HT microplate reader (Biotek). In parallel, cell viability was determined using CellTiter-Glo Luminescent Cell Viability Assay kit (Promega).
Immunoprecipitation and Western blotting
Cells were stimulated for the times indicated followed by lysis in M-PER buffer (Pierce) or RIPA buffer (150 mM NaCl, 50 mM Tris/HCl, pH 7.5, 500 μM EDTA, 1.0% Triton X-100 and 1% sodium deoxycholate) containing protease inhibitor cocktail (Pierce) and phosphatase inhibitor cocktail (Sigma). Protein concentration was determined using the Bradford assay (Bio-Rad Laboratories). For IL-1β detection, cells treated with the indicated amounts of glycyrrhizin were stimulated with 10 nM CpG-DNA overnight. Cells were lysed in RIPA buffer, and equal amounts of cell lysates (1 mg of protein) were incubated with goat anti-(mouse IL-1β) antibody (R&D) followed by precipitation with Protein G–Sepharose beads (GE Healthcare). Beads were washed three times in lysis buffer, followed by elution with SDS sample buffer by boiling at 90 °C for 5 min. For Western blot analysis, equal amounts of cell lysate (10–30 μg) or immunoprecipitation sample were loaded on to a 4–12% or 12% Bis-Tris gel (Invitrogen) and transferred on to PVDF membranes (Invitrogen). Antibodies against mouse p-ERK1/2 (ERK is extracellular-signal-regulated kinase) (Thr202/Tyr204), ERK1/2, p-p38 (Thr180/Tyr182), p38, p-SAPK (stress-activated protein kinase)/JNK (c-Jun N-terminal kinase) (Thr183/Tyr185), SAPK/JNK and p-IκBα (inhibitor of NF-κB α) and IκBα were obtained from Cell Signaling Technologies, anti-(mouse IL-1β) was from R&D, antibodies against mouse Fas, COX-2 and tubulin were obtained from Abcam. HRP (horseradish peroxidase)-conjugated secondary antibodies were purchased from GE Healthcare. Blots were developed using ECL® (enhanced chemiluminescence) substrate (GE Healthcare) on a BioSpectrum Imaging System using a BioChemi HR camera (UVP).
RAW 264.7 cells were incubated for 60 min with 40 nM FITC-labelled CpG-ODN (InvivoGen) with or without glycyrrhizin, washed twice in PBS, fixed with 4% (w/v) paraformaldehyde for 10 min and analysed by flow cytometry using a FACScalibur (Becton Dickinson). A minimum of 20000 cells per sample were acquired.
RAW 264.7 cells were pre-incubated for 2 h with 2.5 mM glycyrrhizin before stimulation with 2 μg/ml LPS. Upon stimulation for 60 min, cells were stained with rat anti-TLR4 antibody (Sa15-21 ) and FITC-conjugated anti-rat antibody and analysed directly using flow cytometry.
Glycyrrhizin causes changes in gene expression upon CpG-DNA stimulation
Glycyrrhizin is well known for its anti-inflammatory actions in vivo and in vitro. Its effect on inflammatory responses induced by the activation of specific TLRs, however, has not been studied in detail so far. Thus we aimed to analyse the effects of glycyrrhizin on gene-expression patterns in response to CpG-DNA by microarray analysis. RAW 264.7 cells were stimulated with 10 nM CpG-DNA in the presence or absence of glycyrrhizin. At 4 h after stimulation, RNA was isolated, labelled and hybridized to a mouse oligonucleotide array. The resulting data were normalized as described in the Experimental section and analysed using Genespring software. To compensate for multiple comparisons, the BH procedure was used for P value adjustment. A total of 20 genes were identified that were down-regulated more than 2-fold with a corrected P value of <0.01. A list of these genes is shown in Table 1. The genes that were down-regulated strongest by glycyrrhizin were the pro-inflammatory cytokines Il1a (17-fold) and Il1b (13-fold). Other strongly regulated genes included the chemokines Ccl5/Rantes, Ccl4/Mip1b, Cxcl2/Mip2a and Ccl3/Mip1a. In addition, some genes that are involved in the regulation of signalling were found to be down-regulated significantly (Cish and Nfkbiz). Using the same criteria (BH-adjusted P value <0.01), eight genes were found to be up-regulated by glycyrrhizin more than 1.5-fold. However, these effects were very mild, with the highest induction being 3-fold (see Supplementary Table S1 at http://www.BiochemJ.org/bj/421/bj4210473add.htm). Glycyrrhizin treatment alone did not cause significant changes in gene regulation in unstimulated cells (results not shown).
Strong down-regulation of pro-inflammatory responses by glycyrrhizin
To analyse further the effect of glycyrrhizin on the gene-expression pattern of CpG-DNA activated cells, we performed GO analysis. For this analysis, we did not adjust for multiple comparisons in order to obtain a more extensive gene list. The input list consisted of 103 genes that were down-regulated and 14 genes that were up-regulated more than 2-fold (see Supplementary Tables S2 and S3 at http://www.BiochemJ.org/bj/421/bj4210473add.htm) with a P value of <0.01. As shown in Table 2, highly significant gene enrichment could be observed for the categories ‘immune response’, ‘immune system process’, ‘cytokine activity’, ‘defence response’ and receptor binding. A total of 24 and 2 genes fell into the categories of ‘immune response’ and ‘immune system process’ respectively. In addition, GO terms including ‘inflammatory response’, ‘receptor binding’, ‘response to wounding’, ‘taxis’ and ‘response to stimulus’ were found to be enriched with high statistical significance. Taken together, these data indicate that glycyrrhizin broadly alters the pro-inflammatory response induced by CpG-DNA.
Confirmation of the anti-inflammatory activity on RNA level
To verify the results obtained by microarray analysis, expression of several inflammation-associated genes that showed strong down-regulation upon treatment with glycyrrhizin were individually analysed by QuantiGene. Therefore cells were stimulated with CpG-DNA and expression of selected genes involved in inflammation (Il1b, Ccl5/Rantes, Ptgs2/Cox2, Tnfa, Il1a, Fas, Ccl4/Mip1b) was analysed at different time points. The most striking result was obtained for Il1b, which was induced approx. 12-fold 4 h and 48-fold 8 h after stimulation, whereas barely any induction was observed in cells treated with glycyrrhizin (Figure 1). Ccl5/Rantes was induced approx. 20-fold 4 h, and as much as 40-fold 8 h after stimulation. In glycyrrhizin-treated cells, Ccl5/Rantes expression remained low, with a maximum of only 4-fold increase compared with unstimulated cells. CpG-DNA caused a 6-fold induction of Ptgs2/Cox2 4 h after stimulation, which increased to 12-fold 8 h after stimulation. In contrast, in cells treated with glycyrrhizin, the induction of Ptgs2/Cox2 was significantly attenuated, with only a 2-fold increase 8 h after addition of CpG-DNA. The effect of CpG-DNA on the induction of Tnfa, Il1a, Fas and Ccl4/Mip1b showed similar effects. Tnfa and Il1a expression was induced 5–6-fold 2 h after stimulation with CpG-DNA and stayed high until 8 h after stimulation, whereas no Tnfa and Il1a mRNA was detected in cells treated with glycyrrhizin. Mild Fas induction was observed upon stimulation with CpG-DNA, which was not detected in cells treated with glycyrrhizin. Ccl4/Mip1b expression was only weakly induced 2 and 4 h after stimulation, but expression increased to 7-fold after 8 h. In the presence of glycyrrhizin, Ccl4/Mip1b mRNA levels remained low. These findings are consistent with the microarray data and confirm that numerous inflammation-associated genes are down-regulated by glycyrrhizin.
Glycyrrhizin inhibits the production of pro-inflammatory proteins
To confirm further the anti-inflammatory properties of glycyrrhizin, its effect on the production and release of cytokines and chemokines that were shown to be attenuated by glycyrrhizin at the mRNA level was analysed. MIP-1β, TNFα and RANTES were induced by CpG-DNA, and glycyrrhizin decreased their release in a dose-dependent manner (Figures 2A–2C). Despite considerable efforts, we were unable to detect IL-1β in the supernatant of cell cultures. However, CpG-DNA induced significant intracellular expression of IL-1β which was decreased in cells treated with increasing amounts of glycyrrhizin (Figure 2D, top panel). In addition, Western blot analysis showed that CpG-DNA induced intracellular expression of COX-2 and Fas which was again inhibited by glycyrrhizin in a dose-dependent manner (Figure 2D, bottom panels). These data demonstrate further that glycyrrhizin down-regulates the expression of pro-inflammatory proteins upon stimulation with CpG-DNA.
Glycyrrhizin inhibits pro-inflammatory responses induced by TLR4 ligands
To evaluate whether glycyrrhizin specifically blocks CpG-DNA-induced inflammatory responses, we investigated mRNA and protein secretion of several pro-inflammatory mediators upon stimulation with LPS, a TLR4 agonist that in contrast to TLR9 signals both, in a MyD88 (myeloid differentiation factor 88)-dependent and -independent manner. LPS triggered strong induction of Ccl5/Rantes, Il1b, Ptgs2/Cox2 and Ccl4/Mip1b and weaker induction of Tnfa mRNA levels (see Supplementary Figure S1 at http://www.BiochemJ.org/bj/421/bj4210473add.htm). In glycyrrhizin-treated cells, induction of all these pro-inflammatory genes was almost completely suppressed. Similar results were obtained on protein level (Figure 3A). Glycyrrhizin decreased LPS-induced secretion of TNFα, RANTES and IFNβ by 70–80%. As observed for CpG-DNA, the decrease was dose-dependent (results not shown). Thus glycyrrhizin was able to inhibit pro-inflammatory responses induced by LPS. Unlike RANTES and IFNβ, the induction of TNFα is dependent on MyD88, suggesting that glycyrrhizin blocks inflammatory responses in a pathway-independent manner.
Glycyrrhizin inhibits responses induced by TLR3
To confirm further the broad and pathway-independent anti-inflammatory properties of glycyrrhizin, we tested its ability to block responses induced by a TLR3 agonist, which, in contrast with TLR4 and TLR9, signals solely in a MyD88-independent manner. Poly(I:C)-induced IL-6 and IFNβ secretion was strongly decreased by glycyrrhizin (Figure 3B). Again, the decrease was dose-dependent (results not shown). Taken together, these data demonstrate that glycyrrhizin indeed has a broad inhibitory effect on the cellular responses to several distinct TLR agonists, implying that it interferes with a target common in all TLR pathways.
Glycyrrhizin inhibits NF-κB activation induced by CpG-DNA and LPS
To deepen our understanding of the molecular mechanism by which glycyrrhizin exerts its anti-inflammatory effects, we analysed the signalling pathways activated by TLR9 and TLR4. Since the effect of glycyrrhizin could readily be explained by a block in the NF-κB pathway, which is activated by all TLRs, we first analysed the effect of glycyrrhizin on the activation of NF-κB by measuring the phosphorylation (hence activation) and expression levels of IκBα. As shown in Figure 4(A), left-hand panel, low levels of IκBα phosphorylation could be detected at 30 and 60 min after CpG-DNA stimulation which peaked at 120 min. In the presence of glycyrrhizin, IκBα phosphorylation was appreciably decreased compared with untreated cells. Total IκBα expression levels correlated with these data. Compared with unstimulated cells, protein levels were weakly decreased at 30 min after stimulation, and IκBα further declined up to 60 min after stimulation in untreated cells, indicating NF-κB activation. At 120 min and 240 min after stimulation, IκBα expression increased again. In cells treated with glycyrrhizin, IκBα expression was considerably higher during the whole course of the experiment, with the most pronounced difference detected at 60 and 120 min after stimulation. The effect of glycyrrhizin on the activation of NF-κB appears to be less pronounced compared with gene-expression data; however, it should be mentioned that the amount of CpG-DNA used in experiments described above (10 nM) was not sufficient to induce detectable levels of activation. Thus these experiments were conducted using 100 nM CpG-DNA, a concentration at which glycyrrhizin overall shows considerably weaker anti-inflammatory activities (results not shown).
Similar, yet more pronounced, results were obtained upon stimulation with LPS (Figure 4A, right-hand panel). At 5 min after LPS stimulation, phosphorylation of IκBα could be observed in untreated cells, whereas in glycyrrhizin-treated cells, barely any activation of IκBα could be detected up to 60 min. In agreement with this, IκBα remained stable until 30 min after stimulation in glycyrrhizin-treated cells, whereas it was almost completely degraded 30 min after stimulation in untreated cells. These data indicated that glycyrrhizin decreases phosphorylation and subsequent degradation of IκBα, thereby reducing downstream activation of NF-κB, both upon stimulation with CpG-DNA and LPS.
Glycyrrhizin decreases CpG-DNA- and LPS-induced activation of MAPKs
Besides the activation of NF-κB, TLR signalling also involves the activation of MAPK pathways. To evaluate the specificity of glycyrrhizin for the NF-κB pathway, we next examined its role in MAPK activation by analysing the phosphorylation of JNK, p38 and ERK1/2. As observed for IκBα activation, no phosphorylation of either MAPK could be detected by stimulation with 10 nM CpG-DNA, thus, again, subsequent experiments were performed with 100 nM CpG-DNA, explaining the mild effects observed with glycyrrhizin. Weak activation of JNK could be detected upon stimulation with 100 nM CpG-DNA (Figure 4B, left-hand panel) at 60 min and stronger at 120 min after stimulation which was mildly decreased in glycyrrhizin-treated cells, whereas total JNK expression was unaffected. Compared with unstimulated cells, enhanced phosphorylation of p38 was detected 30 min and 60 min after CpG-DNA stimulation, whereas the signal declined to background levels after 120 min. In the presence of glycyrrhizin, p38 phosphorylation was significantly weaker at both, 30 and 60 min after stimulation, indicating that glycyrrhizin treatment decreased p38 activation. The expression of total p38 was comparable for all time points and independent of the treatment. Similar results were obtained for ERK1/2. Weak activation was detected 30 min after stimulation, which strongly increased at 60 min after stimulation in untreated cells, whereas only low levels of activation were detectable in cells treated with 1 mM glycyrrhizin 30 and 60 min after CpG-DNA stimulation. The overall ERK1/2 protein levels remained unchanged.
Comparable results were obtained upon stimulation with LPS (Figure 4B, right-hand panel). JNK activation was detectable 30 and 60 min after stimulation in untreated cells, whereas barely any JNK phosphorylation could be observed until 30 min after stimulation and only weak activation 60 min after stimulation in glycyrrhizin-treated cells. Total JNK expression remained unaffected throughout the course of the experiment. Strong activation of p38 was detectable 30 and 60 min after stimulation, which was dramatically decreased in cells treated with glycyrrhizin, whereas treatment had no effect on total p38 levels. Similar results were obtained for ERK1/2 phosphorylation, which was strongest 30 and 60 min after LPS stimulation. In glycyrrhizin-treated cells, activation was decreased throughout the course of the experiment. Total ERK1/2 expression was comparable in treated and untreated cells. Collectively, this analysis indicates that glycyrrhizin does not specifically block the NF-κB pathway, but similarly decreases the activation of MAPKs, strongly suggesting that glycyrrhizin interferes with TLR signalling upstream of MAPK and NF-κB activation. Moreover, given that glycyrrhizin is active against various TLRs that signal through distinct pathways, it is likely to target an upstream event common to all TLR pathways.
Glycyrrhizin decreases cellular uptake/attachment of CpG-DNA
Glycyrrhizin appears to act early upon TLR induction and seems to cause a rather broad block of TLR-mediated responses. We therefore asked whether glycyrrhizin affects the overall stimulation by reducing cellular uptake or attachment of TLR ligands. RAW 264.7 cells were incubated with FITC-labelled CpG-DNA for 60 min, and, after extensive washing, fluorescence was measured by FACS analysis. As shown in Figure 5(A), incubation of cells with CpG-DNA induced a significant fluorescence shift. In cells treated with glycyrrhizin, the shift was decreased dose-dependently. The effect with 1 mM glycyrrhizin was considerably weaker, which is most likely to be due to the increased amount of CpG-DNA (40 nM) used in this experiment. Consistently, under these conditions, glycyrrhizin only shows decreased anti-inflammatory effects, whereas it completely blocked pro-inflammatory responses at a concentration of 2.5 mM glycyrrhizin (results not shown).
TLR4 internalization is attenuated in glycyrrhizin treated cells
Stimulation of cells with LPS is known to trigger TLR4 internalization . To confirm further that glycyrrhizin does indeed block initial responses to various stimuli, we tested its effect on the internalization of TLR4 upon stimulation with LPS. Therefore RAW 264.7 cells were pre-incubated with glycyrrhizin followed by stimulation with LPS. TLR4 internalization was measured upon staining with specific antibodies by FACS. As shown in Figure 5(B), surface TLR4 staining was readily detected in unstimulated cells. LPS stimulation for 60 minutes resulted in a strong decrease of the signal, indicating TLR4 internalization. In contrast, in cells treated with glycyrrhizin, LPS stimulation had no effect on TLR4 surface expression. These data clearly demonstrate that glycyrrhizin interferes with responses to LPS by inhibiting the initiation of TLR4 signalling.
Glycyrrhizin does not affect PMA-induced Tnfa induction
Uptake of CpG-DNA is known to be mediated by scavenger receptors [16–18]. Efficient uptake of pathogens by scavenger receptors requires intact membrane microdomains . Similarly, for efficient TLR signalling, lipid raft formation is critical. Interestingly, it has been reported previously that glycyrrhizin causes altered membrane fluidity . Thus we hypothesized that glycyrrhizin might possess the ability to incorporate into membranes, thereby disrupting the membrane architecture. To test this, we analysed the effect of glycyrrhizin on stimulation with PMA, a receptor- and thus membrane-independent stimulus . RAW 264.7 cells were stimulated with different amounts of PMA with or without glycyrrhizin treatment and mRNA levels of Tnfa were measured 4 h after stimulation. PMA dose-dependently induced Tnfa mRNA (Figure 5C). Strikingly, glycyrrhizin treatment had no effect on Tnfa expression, and remained unaffected under all experimental conditions. These results strongly suggest that glycyrrhizin does indeed cause its broad anti-inflammatory actions by interfering with membrane integrity.
Although the anti-inflammatory properties of glycyrrhizin are widely described, a detailed analysis of its effects on the molecular level has not been conducted to date. By using microarray analysis, we found that glycyrrhizin blocked inflammatory responses induced by CpG-DNA, a TLR9 ligand. The expression of a wide range of cytokines and chemokines was also strongly decreased by glycyrrhizin upon stimulation of TLR4. Moreover, treatment with glycyrrhizin resulted in decreased secretion of IL-6 and IFNβ induced by the TLR3 agonist poly(I:C). These effects were accompanied by a broad block in TLR signalling. Glycyrrhizin not only inhibited the activation of NF-κB, but also the activation of the MAPKs p38, ERK1/2 and JNK induced by both TLR4 and TLR9 ligands. Further analysis revealed that cellular attachment and/or uptake of the TLR9 ligand CpG-DNA was strongly decreased and TLR4 internalization was blocked in the presence of glycyrrhizin. The anti-inflammatory effects of glycyrrhizin were specific for receptor-mediated stimuli, as glycyrrhizin did not show any effect on Tnfa induction triggered by PMA, which signals in a receptor-independent manner. These findings indicate that the anti-inflammatory actions of glycyrrhizin are mediated by an interaction with the cell surface rather than by a specific block in the downstream signalling cascades.
Comparison of gene-expression patterns of untreated and glycyrrhizin-treated cells demonstrated a strong down-regulation of numerous pro-inflammatory genes 4 h after stimulation with the TLR9 agonist CpG-DNA. Strikingly, GO analysis of microarray data revealed that a subset of 24 and 25 glycyrrhizin-regulated genes were over-represented in the categories ‘immune response’ and ‘immune system process’ respectively. Other categories that were found to be overrepresented with high statistical significance included ‘cytokine activity’ and ‘inflammatory response’. These data imply that glycyrrhizin causes a rather broad block to CpG-DNA-induced responses.
Notably, glycyrrhizin efficiently blocked the expression of a wide range of pro-inflammatory mediators at very early times after stimulation. These data strongly indicate that glycyrrhizin acts early upon stimulation, by blocking TLR signalling directly rather than by down-regulating one or a few specific genes, resulting in decreased autocrine/paracrine stimulation. We also detected a decrease in the expression of Ifnb and of numerous IFN-regulated genes. Induction of IFN-inducible genes is not caused by CpG-DNA directly, but is most likely to be a consequence of an autocrine/paracrine stimulation mediated by IFNβ. Thus the decreased expression of IFN-regulated genes in the presence of glycyrrhizin is likely to be a consequence of the initial decreased induction of IFNβ.
The anti-inflammatory actions of glycyrrhizin in response to various TLRs could easily be explained by a block of NF-κB activation. In agreement with this, several reports describing anti-inflammatory actions of glycyrrhizin suggest a specific and direct block of the NF-κB pathway on the basis of data demonstrating decreased IκBα degradation or decreased NF-κB transcriptional activation analysed by EMSA (electrophoretic mobility-shift analysis) or nuclear localization studies of p65/RelA [32,33,40,41]. However, these studies failed to analyse possible effects on parallel pathways such as MAPK activation and events further upstream in the signalling cascade to prove specificity for a direct block of NF-κB activity. Our data, in contrast, clearly demonstrate that glycyrrhizin does not block a specific target in the signalling cascade, but instead causes a broad inhibition not only of NF-κB, but also of MAPK activation upon stimulation of TLR4 and TLR9. Thus these data are suggestive of the hypothesis that glycyrrhizin blocks inflammatory responses either by rather unspecifically targeting several distinct molecules within the different TLR pathways, or more likely, by directly interacting with TLRs or, even more globally, by interfering with cell membrane-associated mechanisms.
The hypothesis that glycyrrhizin interferes with membrane-dependent processes is supported by our data that demonstrate that glycyrrhizin interferes with cellular attachment and/or uptake of CpG-DNA providing an explanation for the broad early inhibitory effects of glycyrrhizin. Treatment with glycyrrhizin also inhibited TLR4 internalization, a process for which lipid raft formation is critical. These observations lead to the hypothesis that glycyrrhizin might be incorporated into lipid bilayers, thereby altering the integrity of the plasma membrane. Indeed, Harada  and Harada et al.  demonstrated decreased membrane fluidity in the presence of glycyrrhizin causing the suppression of infectivity of HIV-1, influenza virus and vesicular stomatitis virus. As a consequence to changes in membrane integrity, it could be hypothesized that glycyrrhizin impedes receptor dimerization, complex formation or ligand-receptor binding, resulting in decreased signal transduction. CpG-DNA uptake depends on scavenger receptors which have been described as being membrane-dependent [16–19]. Decreased membrane fluidity or disruption of membrane microdomains could therefore interfere with scavenger receptor formation and might thus account for the observed decrease of CpG-DNA surface binding in the presence of glycyrrhizin. Interestingly, intact membrane microdomains are also critical for the initiation of TLR4 signalling. Its disruption has been shown to interfere with the formation of the functional complex of TLR4–MD2 with CD14, resulting in a block in LPS signalling . Altered membrane integrity triggered by glycyrrhizin could therefore provide an explanation not only for the inhibition of TLR9 signalling, but also for its anti-inflammatory actions upon TLR4 activation. The fact that glycyrrhizin has no effect on Tnfa expression induced by PMA, a stimulus that is known to induce signalling in a membrane-independent manner  supports this hypothesis even further.
In addition to its block of anti-inflammatory responses caused by TLR agonists, glycyrrhizin similarly blocks TNFα release induced by various additional stimuli, including the pro-inflammatory cytokine HMGB1 (high-mobility group B1) and other RAGE (receptor for advanced glycation end-products) ligands such as S100, S100b and AGE (advanced glycation end-product)–BSA as well as antiviral responses induced by type I IFNs (results not shown). Strikingly, signalling induced by these stimuli has also been demonstrated to depend on intact membrane microdomains [47,48]. Since all of these stimuli act through distinct signalling cascades, yet are dependent on an intact membrane integrity, it is likely that the mode of action is indeed mediated by the ability of glycyrrhizin to attach to or incorporate into the plasma membrane thereby attenuating various membrane-dependent processes including initiation of receptor signalling. Interestingly, glycyrrhizin does not show any toxicity, even if cells are treated with very high concentrations (up to 5 mM). In the case of purely non-specific ‘coating’ of the cell membrane, one would expect cellular maintenance and growth as well as intercellular communication being severely impaired, which would be accompanied by a significant decrease in cell viability. However, cell integrity was fully maintained throughout all of our studies, suggesting that glycyrrhizin targets specific microdomains of the plasma membrane rather than interfering with the entire lipid bilayer.
It should be noted that rather high concentrations of glycyrrhizin were required in our experiments to exert anti-inflammatory effects; one reason for this might be the rather hydrophilic character of glycyrrhizin limiting its interaction with the cellular system. So, at present, it is not clear whether such doses can be reached in vivo, especially in the case of systemic administration. Sufficient doses will probably require a local (e.g. topical) application of a suitable glycyrrhizin formulation. On the other hand, replacing polar residues with hydrophobic groups in the molecule could be a promising approach, which is, however, beyond the scope of the present study.
In conclusion, the present study clearly indicates a broad inhibitory effect of glycyrrhizin on pro-inflammatory responses induced by various TLRs in RAW 264.7 macrophages that appears to be mediated by interactions with the cell membrane (such as alteration of membrane microdomains) rather than by a block of a specific intracellular targets, as has been suggested in the literature to date. These data highlight further the importance of membrane integrity for the initiation of efficient innate immune responses. Interestingly, altered plasma membrane architecture has also been reported to be associated with several diseases, including SLE (systemic lupus erythematosus) and RA (rheumatoid arthritis), by inducing hyperactivation of cell signalling . As glycyrrhizin appears to alter the membrane integrity, thereby attenuating signalling, it might be a useful therapeutic not only for the treatment of a broad variety of chronic inflammatory diseases, but also for the treatment of autoimmune disorders.
Bärbel Schröfelbauer was the leading scientist, who conceived and designed the study, acquired (in part), analysed and interpreted the resulting data, drafted the article and had final approval. Johanna Raffetseder planned and conducted LPS stimulation studies, acquired, analysed and interpreted the resulting data, drafted the article and had final approval. Maria Hauner planned and conducted CpG-DNA stimulation studies, acquired, analysed and interpreted the resulting data, drafted the article and had final approval. Andrea Wolkerstorfer conceived and designed experiments involving TLR3, acquired (in part), analysed and interpreted the resulting data, drafted the article and had final approval. Wolfgang Ernst conceived and designed microarray experiments, acquired (in part), analysed and interpreted the resulting data, drafted the article and had final approval, and Oliver Szolar conceived and designed the study, interpreted data, and drafted and revised the manuscript
This work was supported in part by the Austrian Research Promotion Agency and The ZIT Center for Innovation and Technology.
We thank Professor K. Miyake (Institute for Medical Sciences, University of Tokyo, Tokyo, Japan) for providing the Sa15-21 antibody.
Abbreviations: BH, Benjamini–Hochberg; COX, cyclo-oxygenase; DMEM, Dulbecco's modified Eagle's medium; dsRNA, double-stranded RNA; ERK, extracellular-signal-regulated kinase; FBS, fetal bovine serum; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GO, Gene Ontology; IFN, interferon; IκBα, inhibitor of nuclear factor κB α; IL, interleukin; JNK, c-Jun N-terminal kinase; LPS, lipopolysaccharide; MAPK, mitogen-activated protein kinase; MD2, myeloid differentiation factor 2; MyD88, myeloid differentiation factor 88; MIP-1β, macrophage inflammatory protein 1β; NF-κB, nuclear factor κB; ODN, oligodeoxynucleotide; poly(I:C), polyriboinosinic:polyribocytidylic acid; RANTES, regulated upon activation, normal T-cell expressed and secreted; SAPK, stress-activated protein kinase; TLR, Toll-like receptor; TNFα, tumour necrosis factor α
- © The Authors Journal compilation © 2009 Biochemical Society